Apparently they're not going by head to head in cases where one team has a bad loss or bad scare in another game.
I could handle this - if they were consistent.
In fact, an argument FOR Texas over Oklahoma CAN be made. I'll make it:
"The reason we have Texas over Oklahoma is because:
a) they have the most impressive road win of the season (double digits over Alabama)
b) they handily beat the team that inflicted the loss on OU
c) they basically won a game that came down to the last possession and saw Texas commit twice as many penalties and turn the ball over three times to none"
That is a REASONABLE argument, whether one agrees with the points or not.
In the case of Texas/Oklahoma OU almost lost to UCF at home and then did lose to Kansas who was playing a backup QB.
Other good points to supplement the argument.
My problem?
They don't apply this consistently.
See the 2014 Baylor-TCU mess, where TCU was the higher rated team until the final poll - and THEN those idiots went with, "Baylor beat TCU head to head," as if that didn't happen earlier.
In the case of Oklahoma State/Kansas OSU had a bad loss at home to South Alabama while Kansas lost to Texas at Austin and Oklahoma State at Stillwater so no real shame there.
But again - how does "you lost to a different team and had a close game" REALLY translate into "we will treat your win in the polls as a LOSS?"
I know it's a sore spot with me, but I saw this with the 1993 poll, when Notre Dame beat Florida State and then lost on a last play FG at home the next week to Boston College. Amazingly, that loss to BC turned the win over FSU into a loss poll-wise...only the same logic didn't apply when unbeaten WVA beat BC.
This happens all the time, and that's my problem with it.
They can - and do -literally say anything.
They may as well say, "We ranked Ohio State over UGA because, well, we hate UGA."
I mean, that would be as logical as some of their excuses.